Automatic Tuning Algorithm Research Articles

Exploring and forecasting of solar radiation with machine learning methods

Machine learning has recently advanced to the point that a wide range of solar predicting works have been produced. Specifically, one of the most widely used methods at the moment for hourly solar forecasting is machine learning. However, it appears that there is a misconception regarding forecast accuracy—almost all study articles assert to be better than others. However, it appears that there is a misconception regarding forecast accuracy—almost all study articles assert to be better than others. It is illogical for solar forecasters to place their initial wager on a single model for any new forecasting project. Only commercially available versions of these algorithms are employed, and no hybrid models are taken into account to guarantee an equitable comparison. Additionally, the package's automatic tuning algorithm is used to train each model. Overall results show that tree-based methods consistently yield good results. Reliable generation forecasts are becoming more and more necessary for grid operation as distributed renewable power grows in penetration. In order to produce the most accurate day-ahead hourly irradiance forecasts, the current work combines cutting edge Weather Research and Forecasting (WRF) model implementations with machine learning techniques.

Welding Spark Detection on Construction Sites Using Contour Detection with Automatic Parameter Tuning and Deep-Learning-Based Filters.

One of the primary causes of fires at construction sites is welding sparks. Fire detection systems utilizing computer vision technology offer a unique opportunity to monitor fires in construction sites. However, little effort has been made to date in regard to real-time tracking of small sparks that can lead to major fires at construction sites. In this study, a novel method is proposed to detect welding sparks in real-time contour detection with deep learning parameter tuning. An automatic parameter tuning algorithm employing a convolutional neural network was developed to identify the optimum hue saturation value. Additional filtering methods regarding the non-welding zone and a contour area-based filter were also newly developed to enhance the accuracy of welding spark prediction. The method was evaluated using 230 welding spark images and 104 videos. The results obtained from the welding images indicate that the suggested model for detecting welding sparks achieves a precision of 74.45% and a recall of 63.50% when noise images, such as flashing and reflection light, were removed from the dataset. Furthermore, our findings demonstrate that the proposed model is effective in capturing the number of welding sparks in the video dataset, with a 95.2% accuracy in detecting the moment when the number of welding sparks reaches its peak. These results highlight the potential of automated welding spark detection to enhance fire surveillance at construction sites.

Automated Graph Neural Networks Accelerate the Screening of Optoelectronic Properties of Metal–Organic Frameworks

The numerous organic and inorganic components of metal-organic framework (MOF) materials provide intriguing optoelectronic properties. Accurately predicting the electronic structural properties of MOFs has become the main focus. This work establishes two graph neural network models, crystal graph convolutional neural networks and a materials graph network, for predicting the band gaps of more than 10 000 MOF structures and promotes to improve the prediction accuracy through automatic hyperparameter tuning algorithms. Subsequently, for exploring machine learning-assisted screening of MOFs for the broader electronic properties, the screened copper-based MOFs are compared with lead-based MAPbI3 solar cells with respect to the band gaps, densities of states, and charge density distributions, and the results have demonstrated that the overlap of the wave functions between the initial and final states of MOFs is weakened, which is conducive to the improvement of photoelectric performance. The chlorine doping strategy further enhances the advantage. The tuning of the machine learning model and hyperparameters and the doping strategy of halogen elements furnish empirical rules for the design of MOFs with excellent optoelectronic properties.

A Novel Framework to Facilitate User Preferred Tuning for a Robotic Knee Prosthesis.

The tuning of robotic prosthesis control is essential to provide personalized assistance to individual prosthesis users. Emerging automatic tuning algorithms have shown promise to ease the device personalization procedure. However, very few automatic tuning algorithms consider the user preference as the tuning goal, which may limit the adoptability of the robotic prosthesis. In this study, we propose and evaluate a novel prosthesis control tuning framework for a robotic knee prosthesis, which could enable user preferred robot behavior in the device tuning process. The framework consists of 1) a User-Controlled Interface that allows the user to select their preferred knee kinematics in gait and 2) a reinforcement learning-based algorithm for tuning high-dimension prosthesis control parameters to meet the desired knee kinematics. We evaluated the performance of the framework along with usability of the developed user interface. In addition, we used the developed framework to investigate whether amputee users can exhibit a preference between different profiles during walking and whether they can differentiate between their preferred profile and other profiles when blinded. The results showed effectiveness of our developed framework in tuning 12 robotic knee prosthesis control parameters while meeting the user-selected knee kinematics. A blinded comparative study showed that users can accurately and consistently identify their preferred prosthetic control knee profile. Further, we preliminarily examined gait biomechanics of the prosthesis users when walking with different prosthesis control and did not find clear difference between walking with preferred prosthesis control and when walking with normative gait control parameters. This study may inform future translation of this novel prosthesis tuning framework for home or clinical use.

Closed-loop optimal and automatic tuning of pulse amplitude and width in EMG-guided controllable transcranial magnetic stimulation.

This paper proposes an efficient algorithm for automatic and optimal tuning of pulse amplitude and width for sequential parameter estimation (SPE) of the neural membrane time constant and input-output (IO) curve parameters in closed-loop electromyography-guided (EMG-guided) controllable transcranial magnetic stimulation (cTMS). The proposed SPE is performed by administering a train of optimally tuned TMS pulses and updating the estimations until a stopping rule is satisfied or the maximum number of pulses is reached. The pulse amplitude is computed by the Fisher information maximization. The pulse width is chosen by maximizing a normalized depolarization factor, which is defined to separate the optimization and tuning of the pulse amplitude and width. The normalized depolarization factor maximization identifies the critical pulse width, which is an important parameter in the identifiability analysis, without any prior neurophysiological or anatomical knowledge of the neural membrane. The effectiveness of the proposed algorithm is evaluated through simulation. The results confirm satisfactory estimation of the membrane time constant and IO curve parameters for the simulation case. By defining the stopping rule based on the satisfaction of the convergence criterion with tolerance of 0.01 for 5 consecutive times for all parameters, the IO curve parameters are estimated with 52 TMS pulses, with absolute relative estimation errors (AREs) of less than 7%. The membrane time constant is estimated with 0.67% ARE, and the pulse width value tends to the critical pulse width with 0.16% ARE with 52 TMS pulses. The results confirm that the pulse width and amplitude can be tuned optimally and automatically to estimate the membrane time constant and IO curve parameters in real-time with closed-loop EMG-guided cTMS.

Modeling the control processes of the size population of microorganisms using modulated microwave

The paper describes a mathematical model of a control system for the size population of microorganisms using modulated microwave radiation of non-thermal power.The model is implemented in the dynamic environment MATLAB+Simulink, the areas of optimal values of the impact parameters are found. An algorithm for automatic tuning of the modulation frequency during the experiment has been developed.The efficiency of the algorithmis shown with the help of computer simulation.

Contact force estimation for serial manipulator based on weighted moving average with variable span and standard Kalman filter with automatic tuning

Sensorless contact force estimation methods facilitate the application of the serial manipulators to manufacturing as they enable robots to interact with unexpected collisions at low cost. In this paper, an external force estimation approach with no embedded sensors is proposed. The approach combines a Weighted Moving Average (WMA) with variable span, the standard Kalman filter (SKF), and its tuning routines. Improved confidence in the motor output torque is achieved due to the reduction of the measurement noise in the motor current by the WMA. The span of the filter adapts continuously to achieve optimal tradeoff between response time and precision of estimation in real time. With the comprehensive information of uncertainty in motor current noise and measurement errors of individual joints speed, an automatic tuning algorithm of the SKF is presented. Validation of the presented estimation approach in terms of estimation accuracy and response time was conducted on the Universal Robot 5 manipulator with differing end effector loads. It was found that the combined force estimation method leads to a reduction of the root-mean-square error and response time by 55.2% and 20.8% in comparison with the established method. The proposed method can be applied to any robotic manipulators as long as the motor information (current, joint position, and joint velocities) is available. Consequently, the cost of collision recognition could be reduced dramatically.

A tuning algorithm for a sliding mode controller of buildings with ATMD

This paper proposes an automatic tuning algorithm for a sliding mode controller (SMC) based on the Ackermann’s formula, that attenuates the structural vibrations of a seismically excited building equipped with an Active Tuned Mass Damper (ATMD) mounted on its top floor. The switching gain and sliding surface of the SMC are designed through the proposed tuning algorithm to suppress the structural vibrations by minimizing either the top floor displacement or the control force applied to the ATMD. Moreover, the tuning algorithm selects the SMC parameters to guarantee the following closed-loop characteristics: (1) the transient responses of the structure and the ATMD are sufficiently fast and damped; and (2) the control force, as well as the displacements and velocities of the building and ATMD are within acceptable limits under the frequency band of the earthquake excitation. The proposed SMC shows robustness against the unmodeled dynamics such as the friction of the damper. Experimental results on a reduced scale structure permits demonstrating the efficiency of the tuning algorithm for the SMC, which is compared with the traditional Linear Quadratic Regulator (LQR) and with the Optimal Sliding Mode Controller (OSMC).

Improved adaptive impedance matching for RF front-end systems of wireless transceivers

In this paper an automatic adaptive antenna impedance tuning algorithm is presented that is based on quantum inspired genetic optimization technique. The proposed automatic quantum genetic algorithm (AQGA) is used to find the optimum solution for a low-pass passive T-impedance matching LC-network inserted between an RF transceiver and its antenna. Results of the AQGA tuning method are presented for applications across 1.4 to 5 GHz (satellite services, LTE networks, radar systems, and WiFi bands). Compared to existing genetic algorithm-based tuning techniques the proposed algorithm converges much faster to provide a solution. At 1.4, 2.3, 3.4, 4.0, and 5.0 GHz bands the proposed AQGA is on average 75%, 49.2%, 64.9%, 54.7%, and 52.5% faster than conventional genetic algorithms, respectively. The results reveal the proposed AQGA is feasible for real-time application in RF-front-end systems.

A Novel Oppositional Chaotic Flower Pollination Optimization Algorithm for Automatic Tuning of Hadoop Configuration Parameters.

At present, due to the introduction of the big data era, numerous numbers of data are generated consistently. Many applications utilize big data platforms, namely Spark, Hadoop, Amazon web services, and so on, since these platforms use several parameters for tuning that further enhance the operating performances. It requires a long duration of time to tune the parameters because of the complex relationship and large quantity of parameters. As a result, the building of such parameters and performance optimization at a particular duration of time becomes a challenging task. Several auto-tuning approaches are developed to achieve an optimal design. However, these approaches increase the computation time and minimize the efficiency of the cluster. It is necessary to tune the parameters automatically with low computational and processing time as well as to improve the performance of the system. In this proposed approach, a novel automatic parameter tuning system named as Opt. Tuner is proposed to select the Hadoop configuration parameters with less computational time. The optimization of the proposed approach is achieved by the Flower Pollination Algorithm. Here, a chaotic mapping along with Opposition-Based Learning is introduced for population initialization to form a novel Oppositional Chaotic Flower Pollination Algorithm. The main motive of this initialization phase involves in generating better individuals and to guide the search agent more quickly. In this novel approach, 15 configuration parameters are considered for auto-tuning. Finally, the performance of the proposed approach utilizes the wordcount and sort application to investigate the exhibition and proficiency of diverse databases.

Which is the best PID variant for pneumatic soft robots an experimental study

This paper presents an experimental study to compare the performance of model-free control strategies for pneumatic soft robots. Fabricated using soft materials, soft robots have gained much attention in academia and industry during recent years because of their inherent safety in human interaction. However, due to structural flexibility and compliance, mathematical models for these soft robots are nonlinear with an infinite degree of freedom (DOF). Therefore, accurate position (or orientation) control and optimization of their dynamic response remains a challenging task. Most existing soft robots currently employed in industrial and rehabilitation applications use model-free control algorithms such as PID. However, to the best of our knowledge, there has been no systematic study on the comparative performance of model-free control algorithms and their ability to optimize dynamic response, i.e., reduce overshoot and settling time. In this paper, we present comparative performance of several variants of model-free PID-controllers based on extensive experimental results. Additionally, most of the existing work on modelfree control in pneumatic soft-robotic literature use manually tuned parameters, which is a time-consuming, labor-intensive task. We present a heuristic-based coordinate descent algorithm to tune the controller parameter automatically. We presented results for both manual tuning and automatic tuning using the Ziegler-Nichols method and proposed algorithm, respectively. We then used experimental results to statistically demonstrate that the presented automatic tuning algorithm results in high accuracy. The experiment results show that for soft robots, the PID-controller essentially reduces to the PI controller. This behavior was observed in both manual and automatic tuning experiments; we also discussed a rationale for removing the derivative term.

Water Cycle Algorithm Tuned Fuzzy Expert System for Trusted Routing in Smart Grid Communication Network

Finding an optimal route for reliable data delivery in the smart grid communication network (SGCN) is a challenging task due to its dynamic nature. Even though the rule set (RS) and membership function (MF) framed intuitively in our previous fuzzy logic (FL) approach performs the trusted routing satisfactorily, it consumes computational memory and reduces the energy efficiency of the node. To address this issue, in this article, we proposed a novel trust evaluation framework that employs water cycle algorithm (WCA) for automatic tuning of the rule set and membership function for the decision variable to route the packet in an adaptable manner. Variables like distance, link stability, and node honesty are considered for evaluation using WCA in three iterative processes, namely, exploitation, evaporation, and raining to find the near optimal if–then rules and points for the membership functions. An experimental setup is created using Network Simulator-2 (NS2) to evaluate the performance of the proposed trusted routing algorithm in SGCN. Extensive experiments are conducted for three cases, namely, 1) evaluating RS with fixed MF; 2) evaluating MF with fixed RS; and 3) combined evaluation of MF and RS to evaluate the performance of the proposed model. From the simulation results, it is clear that the RS and MF generated by the proposed model is small and compact enough to provide reliable routing in SGCN.

Speedup your analytics

Database and big data analytics systems such as Hadoop and Spark have a large number of configuration parameters that control memory distribution, I/O optimization, parallelism, and compression. Improper parameter settings can cause significant performance degradation and stability issues. However, regular users and even expert administrators struggle to understand and tune them to achieve good performance. In this tutorial, we review existing approaches on automatic parameter tuning for databases, Hadoop, and Spark, which we classify into six categories: rule-based, cost modeling, simulation-based, experiment-driven, machine learning, and adaptive tuning. We describe the foundations of different automatic parameter tuning algorithms and present pros and cons of each approach. We also highlight real-world applications and systems, and identify research challenges for handling cloud services, resource heterogeneity, and real-time analytics.

Automatic hourly solar forecasting using machine learning models

Owing to its recent advance, machine learning has spawned a large collection of solar forecasting works. In particular, machine learning is currently one of the most popular approaches for hourly solar forecasting. Nevertheless, there is evidently a myth on forecast accuracy—virtually all research papers claim superiority over others. Apparently, the “best” model can only be selected with hindsight, i.e., after empirical evaluation. For any new forecasting project, it is irrational for solar forecasters to bet on a single model from the start. In this article, the hourly forecasting performance of 68 machine learning algorithms is evaluated for 3 sky conditions, 7 locations, and 5 climate zones in the continental United States. To ensure a fair comparison, no hybrid model is considered, and only off-the-shelf implementations of these algorithms are used. Moreover, all models are trained using the automatic tuning algorithm available in the R caret package. It is found that tree-based methods consistently perform well in terms of two-year overall results, however, they rarely stand out during daily evaluation. Although no universal model can be found, some preferred ones for each sky and climate condition are advised.

A Variated Monte Carlo Tree Search Algorithm for Automatic Performance Tuning to Achieve Load Scalability in InnoDB Storage Engines

A Variated Monte Carlo Tree Search Algorithm for Automatic Performance Tuning to Achieve Load Scalability in InnoDB Storage Engines

Gain Scheduled Attitude Control of Fixed-Wing UAV With Automatic Controller Tuning

Fixed-wing unmanned aerial vehicles (UAVs) have become increasingly important in military, civil, and scientific sectors. Because of the existing nonlinearities, effective control this type of UAV remains a challenge. This paper proposes a gain scheduled proportional-integral derivative (PID) control system for fixed-wing UAVs where a family of PID cascade control systems is designed for several operating conditions of airspeed. This is done using an automatic tuning algorithm, where the controllers are automatically selected by deploying an airspeed sensor positioned ahead of the aircraft. Furthermore, the actual gain scheduling is carried out by forming an interpolation between the family members of the linear closed-loop system, which ensures a smooth transition from one operating point to another. Experimental results are conducted in a wind tunnel to show the successful design and implementation of the gain scheduled control system for the fixed-wing UAV and the significant performance improvement over a linear control system without controller adaptation.

Разработка алгоритма автоматической подстройки центральной частоты спектроанализатора

Разработка алгоритма автоматической подстройки центральной частоты спектроанализатора

SIFT-FANN: An efficient framework for spatio-spectral fusion of satellite images

Image fusion techniques are widely used for remote sensing data. A special application is for using low resolution multi-spectral image with high resolution panchromatic image to obtain an image having both spectral and spatial information. Alignment of images to be fused is a step prior to image fusion. This is achieved by registering the images. This paper proposes the methods involving Fast Approximate Nearest Neighbor (FANN) for automatic registration of satellite image (reference image) prior to fusion of low spatial resolution multi-spectral QuickBird satellite image (sensed image) with high spatial resolution panchromatic QuickBird satellite image. In the registration steps, Scale Invariant Feature Transform (SIFT) is used to extract key points from both images. The keypoints are then matched using the automatic tuning algorithm, namely, FANN. This algorithm automatically selects the most appropriate indexing algorithm for the dataset. The indexed features are then matched using approximate nearest neighbor. Further, Random Sample Consensus (RanSAC) is used for further filtering to obtain only the inliers and co-register the images. The images are then fused using Intensity Hue Saturation (IHS) transform based technique to obtain a high spatial resolution multi-spectral image. The results show that the quality of fused images obtained using this algorithm is computationally efficient.

Automatic tuning of robust constrained cross‐direction controllers

SummaryThe paper machine cross‐directional (CD) process is a large‐scale spatially distributed system. It is known to be severely ill‐conditioned as the gain rolls down to zero for some of the process directions. Model uncertainties in the process are inevitable resulting in a challenging robust control design problem. CD actuators are subject to min–max constraints while slice lip actuators are subject to additional bending moment limits. Because of the large number of input constraints, the industrial practice is to tune the CD controller assuming inactive constraints. The robustness of CD feedback loops to model uncertainties under constrained internal model control satisfies an integral quadratic inequality. This work develops an automatic tuning algorithm that guarantees robust stability and performance of the constrained CD feedback loop. Spatial response models are identified in a prediction error frame delivering bounds on the CD process pseudo‐singular values. The CD controller is synthesized online through a linear matrix inequalities feasibility problem taking into consideration the modal space uncertainty rising from the uncertainties in the estimated parameters and the expected variations in the dynamic response. The developed tuning technique is suitable for paper machines producing different grades of paper as the CD process spatial and dynamic responses change for each grade. The performance of the tuned constrained internal model control controller is validated through comparing it to an industrial CD controller that has been implemented in paper mills as part of a commercial product. Copyright © 2015 John Wiley & Sons, Ltd.

Automated Two-Degree-of-Freedom Model Predictive Control Tuning

This work considers the automated tuning of a two-degree-of-freedom model predictive controller (MPC) for single-input, single-output industrial processes with model uncertainties. The objective of the tuning algorithm is to automatically determine the MPC tuning parameters such that (1) the robust stability can be guaranteed, (2) the worst-case overshoot is controlled, (3) the oscillations in process outputs are attenuated, and (4) the worst-case settling time is minimized. A rigorous robust stability analysis is first conducted based on the connection between parametric uncertainties and unstructured uncertainties, and a tight robust stability condition is derived. As the specification on process output variation is not easily made by the end users, two alternative methods are proposed to automatically determine the tolerable total variation, which lead to two automatic tuning algorithms that achieve the tuning objectives. The proposed results are tested and verified through examples extracted from industrial processes in the pulp and paper industry, and comparisons are made with other existing results.